Abstract
China’s tobacco plantation industry includes a large scale and numerous employees. Nevertheless, the current tobacco curing control technology has higher labor intensity, however, it also cannot be adjusted based on the conditions of different tobacco leaves batches reducing the quality of tobacco. The tobacco’s quality can be improved after curing and labor intensity can be reduced by modeling the state prediction of the tobacco curing process, accurately predicting the state of the tobacco curing, and making timely adjustments to the curing process. The area, color, weight, and some chemical substances of tobacco leaves significantly change during the tobacco curing process. This can be theoretically used as the input feature of the state prediction model. However, it is difficult to calculate the changes of the area and the changes of chemical substances in real-time owing to the complexity of the intensive curing room environment. Only features such as color and weight are easier to extract indicating that the model has fewer available features, and the accuracy of prediction using a single model is relatively low. Considering this problem, a state prediction fusion model (SPFM) was proposed integrating Long-Short Term Memory (LSTM) and Extreme Gradient Boosting (XGBoost). At the same time, based on the characteristics of the data set, a new data processing is proposed for the tobacco curing data set. By denoising, the image, the characteristic values of the RGB (Red, Green, Blue), HSV (Hue, Saturation, Value) color space were extracted. Then, the data pre-processing such as standardization, data cleaning, and label digitization were performed on the data. Furthermore, an intelligent tobacco curing platform was designed to integrate data collection, online monitoring, data mining, and status prediction, and SPFM was embedded in the platform. A comparative test was conducted, based on the real data collected via a tobacco station in 2019. The results indicate that SPFM has a better performance compared to support vector machines, artificial neural network, and the base models of SPFM such as XGBoost and LSTM. The accuracy of SPFM is 0.974, with an increase of 4.8–59.7%; the macro recall of SPFM is 0.952, with an increase of 8.2%–49.9%; the macro recall of SPFM is 0.936, with an increase of 8.2–75.0%; and the macro F1-score of SPFM is 0.943, with an increase of 9.7–108.6%.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
Chen E, Fan Z, Song Z, Wang D, Hu Y, Wang X et al (2017) An analysis of the law and key influencing factors of tobacco leaf mildew during baking. Acta Agriculturae Universitatis Jiangxiensis 39(5):877–883
Chen T, Guestrin C (2016) Xgboost: A scalable tree boosting system. In Proceedings of the 22nd acm sigkdd international conference on knowledge discovery and data mining, pp 785–794
Cuirong PQY (1999) Research on intelligent temperature humidity controller for tobacco baking [j]. Chin J Sci Instr 3
Eltag K, Aslamx MS, Ullah R (2019) Dynamic stability enhancement using fuzzy pid control technology for power system. Int J Control Autom Syst 17(1):234–242
Gers FA, Schraudolph NN, Schmidhuber J (2002) Learning precise timing with lstm recurrent networks. J Mach Learn Res 3:115–143
Goutte C, Gaussier E (2005) A probabilistic interpretation of precision, recall and f-score, with implication for evaluation. In: European conference on information retrieval, pp 345–359
Graf H, Cosatto E, Bottou L, Dourdanovic I, Vapnik V (2004) Parallel support vector machines: the cascade svm. Adv Neural Inf Process Syst 17:521–528
Hochreiter S, Schmidhuber J (1997) Long short term memory. supervised sequence labelling with recurrent neural networks. Springer Berlin Heidelberg, 1735 , 1780
Hopfield JJ (1982) Neural networks and physical systems with emergent collective computational abilities. Proc Natl Acad Sci 79(8):2554–2558
Jing X, Cheng L (2012) An optimal pid control algorithm for training feedforward neural networks. IEEE Trans Industr Electron 60(6):2273–2283
Kawakami K (2008) Supervised sequence labelling with recurrent neural networks. Ph. D. thesis
Levin A (2019) Three-stage process of baking, infusing, and sealing cakes. Google Patents. (US Patent App. 15/697,302)
Li L, Yu Y, Bai S, Cheng J, Chen X (2018) Towards effective network intrusion detection: a hybrid model integrating gini index and gbdt with pso. J Sensors
Liu J, Shang W, Lin W (2018) Improved stacking model fusion based on weak classifier and word2vec. In: 2018 ieee/acis 17th international conference on computer and information science (icis), pp 820–824
Macdonald C, Ounis I (2006) Voting for candidates: adapting data fusion techniques for an expert search task. In: Proceedings of the 15th acm international conference on information and knowledge management, pp 387–396
Pan Y, Shen P, Shen L (2012) Speech emotion recognition using support vector machine. Int J Smart Home 6(2):101–108
Qazanfari H, Hassanpour H, Qazanfari K (2019) Content-based image retrieval using hsv color space features. Int J Comput Inf Eng 13(10):537–545
Qiu W, Qiu Z (2006) Design for temp-humidity control system of tobacco parching house based on fuzzy-pid control. In: 2006 international conference on mechatronics and automation, pp 2229–2234
Schmidhuber J (2015) Deep learning in neural networks: an overview. Neural Netw 61:85–117
Shanmuganathan S (2016) Artificial neural network modelling: an introduction. Artificial neural network modelling. Springer, New York, pp 1–14
Sharma A (2019) Understanding rgb color spaces for monitors, projectors, and televisions. Inf Display 35(2):17–43
Sheng J, Chun X, Jianhua X, Yonghui Z, Jinchao Z, Qiang X, Yong G (2018) Analysis of disastrous climate and its coping mechanisms and measures in the tobacco-growing areas of luzhou. Meteorol Environ Res 9(4):33–40
Shi X, Chen Z, Wang H, Yeung D-Y, Wong W-K, Woo W-c (2015) Convolutional lstm network: A machine learning approach for precipitation nowcasting. ar Xiv preprint. arXiv:1506.04214
Sural S, Qian G, Pramanik S (2002) Segmentation and histogram generation using the hsv color space for image retrieval. In: Proceedings. international conference on image processing (Vol. 2, pp. II–II)
Süsstrunk S, Buckley R, Swen S (1999) Standard rgb color spaces. In Color and imaging conference, Vol. 1999, pp 127–134
Suykens JA, Vandewalle J (1999) Least squares support vector machine classifiers. Neural Process Lett 9(3):293–300
Wang F, Li Y, Liao F, Yan H (2020) An ensemble learning based prediction strategy for dynamic multi-objective optimization. Appl Soft Comput 96:106592
Wang F, Li Y, Zhou A, Tang K (2019) An estimation of distribution algorithm for mixed-variable newsvendor problems. IEEE Trans Evol Comput 24(3):479–493
Wang Y, Huang M, Zhu X, Zhao L (2016) Attention-based lstm for aspect-level sentiment classification. In: Proceedings of the 2016 conference on empirical methods in natural language processing, pp 606–615
Wu J, Yang SX (2019) Intelligent control of bulk tobacco curing schedule using ls-svm-and anfis-based multi-sensor data fusion approaches. Sensors 19(8):1778
Yu B, Qiu W, Chen C, Ma A, Jiang J, Zhou H, Ma Q (2020) Submito-xgboost: predicting protein submitochondrial localization by fusing multiple feature information and extreme gradient boosting. Bioinformatics 36(4):1074–1081
Zhang B, Wang X et al (2018a) Effects of different temperature and humidity combinations at yellowing and color-fixing stages of baking process on leaf quality of tobacco variety cuibi-1. Acta Agriculturae Jiangxi 30(3):88–92
Zhang Z, Zhang K, Khelifi A (2018b) Multivariate time series analysis in climate and environmental research. Springer, New York
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Wang, Y., Qin, L. Research on state prediction method of tobacco curing process based on model fusion. J Ambient Intell Human Comput 13, 2951–2961 (2022). https://doi.org/10.1007/s12652-021-03129-5
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s12652-021-03129-5